I have seen 21700s in 50G, 50E i think there is a S and a T as well.
Amperage / Capacity
1 Thank
From what I’ve seen, there’s Samsung 30T, 40T, 48G, 48X, 50E, 50G, 50S.
however, there are some “generation” for some of these batteries (I’ve noted based on some readings here in BLF before that there is a 40T3, 50E2, 50E3. (seem to recall 50E2 or was it 50E3 approaches the performance/capacity of 50G).
Again, I’m not an authority but just mentioning from some personal experience with Samsung 21700, and also from what I recall reading here.
3 Thanks
You already know the current delivery vs. capacity thing, but the incremental changes to some cells usually give us less voltage sag under load while maintaining the capacity (or even improving it slightly). So like the 50E vs the 50E3 is just a nice slight improvement for its intended application (drain) but it still isn’t capable on the high end of competing with the higher drain 40T and such.
The 50G is a definite improvement over the 50E in both voltage sag and capacity. 50S widened the gap with higher capacity while also being able to hold up under decent amp draw, but it seems it comes at the high cost of greatly reduced cycle life and faster diminishing capacity after only a couple hundred cycles.
All that said, if you take a situation like stressing a 50E at 10amps or so vs. the same on a 40T, the actual capacity and run time in use like that may be a wash and they’ll be very close to one another. Look at test graphs like HJK and Mooch to get an idea of what is really happening under load, and you can choose the best for your application.
Recently I picked up some of the Vapcell G50 cells (which are rewrapped Samsung 50G cells, but was the only way I could get my hands on that cell here in the US). In most of my lights the 50E is just fine and performs great, but after looking at the lower voltage sag over the entire capacity range of the 50G, I got some to put in UV lights - because UV emitters are voltage hungry and having less sag just helps eke out more light and run time on those. In 18650 UV lights I did the same by using the Sanyo/Panny 18650GA instead of the Samsung 35E and Sony VTC6 and the performance increase for that application is slight but noticeable for how I use the lights.
So basically, shop by the capacity you want, but check to see whether the current delivery also meets your needs, and real tests/graphs are better real information than published specs or advertising. And try to keep your drain around 80% ish of the tested/rated max current drain if you’re interested in less stress on the cells/slightly better service life.
3 Thanks
Is there a chart of cycles and current capacity.
I have googled it in the past but not found any.
Not that I’ve ever seen but that would be handy.
1 Thank
One thing I’ve seen but never understood is a phenomenon where a cell performs better in one light than another, occasionally significantly better. Given the same (brand , type) of emitters and simply testing the cell (fully charged each time) in different lights.
To see some lights do better with a (specific cell) while typically a (different cell) is the top performer is confusing.
This has occurred testing different cells in a light to see which cell was top. Same set of cells (14 different kinds as I recall) showed a different results in another light, no confirmation for the top dawg contest.
I’ve done this quite often in order to be able to recommend a cell by performance. Can be difficult to say.
If none of this makes sense it’s probably exhaustion, been cleaning up ice damage and mowing (ironic, I know)
1 Thank
I would guess that could come down to various resistances, separate or in combination. A little different IR in the cells, variation in the emitter bin, etc…possibly even the spring material (or an extra half coil or long tag end)…oxidized or greasy contact points, too. Other than 7135s do other chips have binning variation like we see in emitters…enough to matter?
Or Murphy.
2 Thanks
Do you REALLY think I’d be testing with oxidized or greasy contacts? Me? The guy looking for absolutes?
1 Thank
Samsung is transitioning away from high output power tool batteries and going all in on high capacity, low output, low cost EV batteries and moving away from cobalt towards nickel and silicon. I would expect most of their cells to start to follow this trend of they haven’t already.
Well, I mean it could be your own blood, or red dirt sweat, or anything. 
What do you think about chips…plausible contribution? I really don’t know and never dig into datasheets for those.
I’ve studied the data sheets on dozens of MOSFETS, back when we were developing the direct drive drivers. I think I found the SIR800 about the time Comfychair did, maybe AlexGT as well. I found the SIR404 for our needs then I think Del discovered the Infineons.
Wight taught me what to look for in the datasheets.
1 Thank
I have a 50T that’s dark grey, not sure what it’s designed for, doesn’t seem to be able to hit hard as compared to the 40T although the IR shows 16 where my 40T’s are 18-19.
The big 16 core SFH55 isn’t impressed with it… or my 40T’s, or my SIR404, or the 18 ga leads and bypasses or the silver plated beryllium spring. Close to 5000 lumens where the chase was aiming between 9000 and 14,000. Another case of Chinese overestimating?
Edit: I discovered that I had my dummy cell a little on the short side and there wasn’t good spring compression. I remedied this with a quick and dirty brass cap on the beryllium spring of the driver for a substantial gain… even the 50T shows 7140 lumens at start up, the 40T shows 7640 briefly then settles at 6900 with a slow drop from there. So it will be interesting finding the right cell for the SFH55.
1 Thank
Some things never change. lol. I haven’t tried any of these new Chinese emitters but have followed along the threads and comments here. It doesn’t seem like many of them have really come up to par with the majors yet but there are some nice ones. Actually Wurkkos used some little E21-sized flats from Latticepower (fka Lattice Bright from back in the “fake days”) and those are quite impressive little things. Just in their TS10 as far as I know.
That’s an LG cell. I think Illumn was selling them. Lower drain like the Samsung 50E.
Looks like an LG M50LT, right on the wrapper. The website on the anti-vaping label on the wrapper is also an LG website.
Should be around 5000mAh and max discharge around 14A, don’t put it in any hotrods as AFAIK they do not handle high currents well.
1 Thank
Good to know, thanks!
Do you have a link on this?
Is it a high performance cell?
I’m told the cell I pictured is an LG, yes it runs the hungry SFH55 almost as well as the 40T.
I’ve also been told it doesn’t fare well under high loads but this one seems to be doing fine, maintaining around 16mOhms internal resistance on my Vapecell charger.
1 Thank